HPLC analysis of estrogen receptor by a multidimensional approach

Expression of estrogen receptor variants in normal and neoplastic human uterus

Estrogen receptor variants lacking internal exons and representing dominant positive and negative activity may be involved in the initiation and/or progression of endocrine dependent tumors. To assess the role of estrogen receptor in uterine disease, we have analyzed both normal and neoplastic uterine samples for the presence of variant estrogen receptors using the sensitive technique of RT-PCR and direct automated DNA sequencing of the amplified products. Our analysis was conducted to determine the presence of spliced variants lacking exons 3 through exon 8. We demonstrate that both the normal and neoplastic human uterus contains a number of spliced variants of the estrogen receptor that co-exist with the wild type receptor. Variants lacking exons 4, 5 and 7 but not exons 3 and 6 were detected. Also, a novel partial deletion in exon 8 was detected in both the normal and neoplastic tissues, although a total deletion of this exon was not observed. In addition another region of exon 8 deletion was found to be present in one tumor tissue which also contained an insertion within this region, however, other tumors did not contain this variant. In addition, double exon deletion variants were observed lacking exons 3 and 4, exons 4 and 5, and exon 7 with part of exon 8. Although our data represents a limited number of samples it suggests that splicing of the estrogen receptor message occurs in the normal physiological setting. There does not appear to be any association between the presence or absence of spliced variants of estrogen receptor and uterine tumor formation at the mRNA level.

Simultaneous identification of estrogen and progesterone receptors by HPLC using a double isotope assay

Polymorphism of estrogen (ER) and progestin receptors (PR) was analyzed simultaneously using high performance hydrophobic interaction chromatography (HPHIC). HPHIC was used previously to characterize four ER isoforms [Hyder et al., J. Chromat. 397 (1987) 251] based on retention times on Synchropak propyl (100 x 6 mm) HPLC columns (Synchrom, Inc.). ER and PR were prepared from human breast cancer. ER was labeled with 3 nM of either [3H]estradiol-17 beta ([3H]E) or [125I]iodoestradiol-17 beta ([125I]E) while PR was associated with 5 nM of either [3H]R5020 ([3H]R) or [125I]iodovinylnortestosterone ([125I]V). ER was resolved by HPHIC into isoforms MI (Rt = 11 min), I(Rt = 16 min), and II (Rt = 24 min). Isoforms I and II each accounted for ca 45% of specific binding. PR separated into isoforms MI (Rt = 14 min) and I (Rt = 21 min, 80% of specific binding) when eluted with the same gradient used for ER chromatography. Upon inclusion of 10 mM molybdate ER resolved into isoforms MI and MII (Rt = 16 min) and PR into isoforms MI and I (here however isoform MI represented 80-95% of specific binding). Elution patterns were preserved with different batches of stationary phase suggesting the integrity of the isoform distribution. HPLC profiles of ER isoforms labeled with earlier [125I]E or [3H]E were identical as were PR isoform profiles labeled with either [3H]R or [125I]V. Pairs of 125I- and 3H-labeled ligands were used in either combination to monitor ER and PR profiles simultaneously. Isoforms analyzed in 50 biopsies gave reproducible retention times, however the ratio between I and II for ER and MI and I for PR varied. This method allows rapid, simultaneous monitoring of the chromatographic behavior of ER and PR isoforms or other associating proteins or nucleotides. One may now better elucidate their interrelationship as it relates to the hormone-response mechanism.

Immunological and physiological analysis of human breast cancer progesterone receptor heterogeneity, following KCl dissociation and size exclusion chromatography

Progesterone receptor (PR) levels were determined in 69 human breast cancer specimens by both radioligand assay (RLA) and enzyme immunoassay (EIA). These methods did not detect the same number of sites, and for each tumor there was a constant ratio between epitopes and PR-binding sites corresponding to 1/4, 2/4, ... 8/4. High performance size exclusion chromatography was performed to separate the various PR isoforms, and the ability of these isoforms to interact with the monoclonal antibodies was assessed. Determination of PR in the chromatographic fractions by EIA and RLA showed that the various isoforms isolated by chromatography presented variable quantities of steroid-binding sites and epitopes, thus confirming the differences observed in the cytosol assays. The dissociation of molybdate-stabilized PR by KCl and measurement by RLA and EIA of the isoforms obtained showed two different types of chromatographic patterns, particularly in the 8S polymeric form where the monoclonal antibodies appeared to detect mainly the heavier 8S fraction, which may correspond to the 8S-B form of the progesterone receptor. The monoclonal antibodies also detected an intermediate PR polymeric form (236 kDa) which was not always detected by the tritiated ligand. Our results suggest that breast cancer PR exhibit a marked molecular heterogeneity which may partially explain the differences in response to hormonal therapy, particularly for tumors with high receptor levels which nonetheless fail to respond to treatment.

Oestrogen receptor activation and molecular forms in human breast cancer

The oestrogen receptor (ER) present in human breast cancer cytosol was resolved by high performance size exclusion chromatography into a high molecular weight form (300 kDa) either alone or together with a lower molecular weight species (60 kDa). In tumours that contained the 60 kDa species there was a significant increase both in ER activation, determined by binding to oligo(dT)-cellulose, and the concentration of progesterone receptor (PR). Despite cytosolic ER being a potential homogenisation artefact, qualitative differences observed in vitro may reflect aspects of tumour biology in vivo.

High-performance hydrophobic interaction chromatography as a means of identifying estrogen receptors expressing different binding domains

Methodology for high-performance hydrophobic interaction chromatography (HPHIC) of estrogen receptors (ER) was developed, utilizing a polyether-bonded stationary phase, which was non-ionic in nature. Using a descending salt gradient (2 M to 0 M ammonium sulphate in 40 min), ERs from human breast cancer separated into two isoforms, which retained ligand-binding domains. The same isoforms were observed with ER preparations from rat uterus. When sodium molybdate, a stabilizer of receptor structure, was incorporated into the mobile phase, it altered the ER characteristics, producing an earlier elution of one component, while the other one remained unchanged. Treatment of breast cancer cytosol with RNase A did not alter ER elution from either the hydrophobic or size-exclusion (TSK 3000 SW) columns. Modification of cysteine residues with N-ethylmaleimide led to a broad elution pattern of receptor from the hydrophobic column, implying the existence of multiple conformations of ER. Limited trypsin treatment of ER, which removes the DNA binding domain, led to the elution of only one receptor peak from the hydrophobic column. The receptor eluted at 24 min both in the presence and in the absence of sodium molybdate. Thus, at least one mechanism of the sodium molybdate effect must involve its direct interaction with ER to influence the sequence between the DNA-binding domain and the N-terminus. This also indicates that the most hydrophobic species of ER (sodium molybdate sensitive) may arise due to the interaction of the DNA-binding site with the stationary phase. Other possibilities, such as differential post-translational modifications of the receptor protein could also account for the two isoforms of ER, observed in HPHIC analysis.

High-resolution separation of molybdate-stabilized progestin receptors using high-performance liquid chromatography

Molecular heterogeneity of the human uterine progestin receptor was investigated employing sucrose density gradient centrifugation and high-performance liquid chromatography (HPLC) in size-exclusion (HPSEC), ion-exchange (HPIEC) and chromatofocusing (HPCF) modes. Synthetic progestomimetic ligands, [3H]R5020 and [3H]ORG-2058, were used to identify these receptors. Rapid centrifugation with a vertical tube rotor showed both 8-9 S and 4-5 S receptor species in the presence of 10 mM sodium molybdate with a 90-96% recovery. [3H]R5020 displayed greater nonspecific binding than [3H]ORG-2058. When separated receptor preparations were labeled, each with a different ligand, mixed and separated on optimized gradients, at least two receptor isoforms were identified in the components sedimenting at 8-9 S. HPSEC confirmed the presence of receptor isoforms displaying different molecular size and shape dependent upon the progestin ligand used. When the surface charge properties were examined by HPIEC using AX-1000, two distinct species were observed irrespective of the radioactive ligand. The first peak appeared in the void volume similar to the position of free steroid, indicating the possibility of ligand stripping by the column. The second peak bound steroid specifically and eluted with 100 mM phosphate. If either 8-S or 4-S progestin receptors were first separated by gradient centrifugation then by HPIEC, both receptor isoforms eluted with 60 mM phosphate. Re-chromatography of these on HPIEC also gave the isoform eluting at 60 mM phosphate. HPCF of ligand-bound receptors on AX-500 columns also identified one isoform eluting at pH 5.6-6.1. Using a combination of HPLC techniques and sucrose gradient centrifugation, heterogeneity of the progestin receptor has been demonstrated.

High-performance hydrophobic interaction chromatography of estrogen receptors and magnesium dependent protein kinase(s): detection of two molecular forms of estrogen receptors in the presence and absence of sodium molybdate

The separation characteristics of estrogen receptors (ER) from human breast cancer were evaluated based on their hydrophobic properties. Results show that (1) two distinct hydrophobic isoforms of ER exist either in the presence of sodium molybdate (peaks MI and MII with retention times of 15-17 min and 24-26 min) or in its absence (peaks I and II with retention times of 25-27 min and 34-36 min respectively); (2) this is observed whether molybdate (MoO2-4) is added to prepared cytosol or to the buffer prior to homogenization; (3) isoform MII and I separated with similar retention times suggesting they are the same ER species; and (4) isoform MI (Rt = 15-17 min) is a distinct ER species from either MII/I (Rt = 25-28 min) or II (Rt = 34-36 min). The latter isoform represents a highly hydrophobic species seen only in the absence of MoO2-4. Finally, (5) MoO2-4 ions appear to interconvert the most hydrophobic species (II) into the least hydrophobic isoform (MI) with virtually no change in the quantity of isoform(s) MII/I. However, it cannot be ascertained if the II----MI interconversion proceeds via isoform MII/I. Isoform II may result from the interaction with the stationary phase via its DNA binding site since MoO2-4, which is suggested to directly interact with this site, selectively interacts with peak II. These results imply the usefulness of inclusion of receptor stabilizing reagents in the mobile phase for preserving receptor integrity and in elucidating the interrelationships of ER isoforms and associated macromolecules.

Investigation of the corticosteroid receptor system in rat hippocampus by ion exchange fast protein liquid chromatography

In order to study the receptor system for adrenocortical steroids, hippocampal cytosolic preparations--containing both type I and type II receptors--were subjected to anion exchange fast protein liquid chromatography (FPLC). With running buffer containing Tris, EDTA, and glycerol three peaks (1-3) were eluted from the column at 220, 400 and 560 mM NaCl respectively regardless of whether [3H]corticosterone or [3H]RU 28362 had been used as radiotracer. None of the peaks was caused by serum transcortin as revealed by control studies. However, the sequestering influence of transcortin on receptor binding of corticosterone could be demonstrated by the FPLC technique with mixtures containing serum and hippocampus cytosol. Competition experiments with cytosolic samples revealed that type I receptor was present only in peaks 2 and 3 while type II was found in all three peaks in variable amounts, depending on the presence of molybdate. When molybdate was added to the running buffer only two peaks (2 and 3) were eluted, both containing type I and type II receptors. Peak 1 was attributed to the activated type II receptor while peak 2 represented nonactivated receptors. The origin of peak 3 remains uncertain. The data indicate that molybdate must be present in the cytosolic preparation and in the running buffer to keep type II receptor in its nonactivated form. Type I receptor was probably not transformed into the activated form in the absence of molybdate but lost binding capacity and/or affinity for corticosterone.

Characterization of estrogen receptors and associated protein kinase activity by high-performance hydrophobic-interaction chromatography

We have determined that high-performance hydrophobic-interaction chromatography (HPHIC) with weakly hydrophobic columns permit the rapid separation of the labile isoforms of estrogen receptor proteins. Previously we reported the use of the SynChrom propyl 500 column for HPHIC of steroid receptors. However, due to the strongly hydrophobic characteristics of the ligand, [125I]iodoestradiol-17 beta, and the receptor protein, organic solvent was required in the mobile phase for greater recovery of receptor proteins. Here, we report separation of steroid receptors from human breast tumors and rat uteri, using the Beckman CAA-HIC, a non-ionic polyether-bonded column, without the need for organic solvents and with virtually 100% recoveries. Receptors were extracted in 10 mM phosphate buffer (pH 7.4). Maximum resolution and separation were achieved when a descending salt gradient of ammonium sulfate in phosphate buffer (pH 7.4) was used (2-0 M in 30 min). Estrogen receptor (ER) was resolved into two isoforms with tR = 22 +/- 1 min (n = 16, designated as peak I) and 27.5 +/- 0.5 min (n = 14, designated as peak II) and a purification of five- to twenty-fold in a single pass. Free steroid was eluted at tR = 35 +/- 1 min (n = 4). Separation was dependent on adjusting the ionic strength of cytosol to 1.5 M ammonium sulfate. ER, purified by HPHIC, retained ligand binding capacity and exhibited protein kinase activity, which was dominant in the less hydrophobic peak I (tR = 22 min) when immunoprecipitated with the monoclonal antibody D547. This method of rapidly purifying ER with high retention of biological activity may now be applied to the study of the molecular interrelationships of steroid receptor isoforms.

The effects of ischemia on estrogen and progesterone receptor profiles in the rodent uterus

Ischemia is considered to invalidate the estrogen receptor (ER) and progesterone receptor (PR) values of tissues available for steroid hormonal analysis. To evaluate the temporal effect of devascularization on steroid receptors, complete uterine ischemia was induced in vivo in 64 female Buffalo rats (80-154 g). Animals were assigned to varying intervals of ischemia (0-90 min) and were maintained at a constant ambient (75 degrees F) and core-body temperature (100-102 degrees F). Following tissue preservation at -80 degrees C, multipoint titration of steroid binding capacity (SBC) was performed with [3H]estradiol 17-beta (3-0.15 nM) or [3H] R5020 (8-0.4 nM) in the presence or absence of a 200-fold excess of an unlabeled ligand. Applying nonlinear regression analysis, ischemia was observed to decrease the binding capacitance for both ER and PR profiles of rodent uterine tissues at 30 min with significant decay over the 90-min interval of devascularization (ER, P less than 0.01; PR, P less than 0.1). Significant reduction in SBC was evident after 80 min of ischemia for PR (P less than 0.05) and 90 min of tissue ischemia for ER (P less than 0.05) comparative to control (t0) valves. The detrimental effect of progressive ischemia on ER and PR values was such that it appears essential to assure rapid and reliable tissue aliquot preservation techniques when organ ischemia greater than or equal to 90 min is anticipated.